Breast cancer that does not respond to estrogen (estrogen receptor negative) comprises 30% of invasive breast cancers. These tumors are more aggressive, tend to metastasize, and have a poor prognosis. Cyclooxygenase (COX)-2 and the epidermal growth factor receptor (EGFR) may be potential targets in breast cancer prevention and treatment. COX 2 is overexpressed in many types of human tumors including breast. The COX-2 inhibitors have been shown to reduce the incidence of colon and mammary tumors in rodents and inhibit the growth of established mammary tumors. Although their mechanism of action is not clear, these drugs have been shown to inhibit the COX enzyme, reduce prostaglandin synthesis, and suppress angiogenesis, which is essential for tumor growth. EGFR is overexpressed in both estrogen receptor positive and estrogen receptor negative human breast cancers and is also associated with a poor prognosis. ZD1839 (Iressa) is a potent antitumor agent that targets EGFR tyrosine kinase and has been shown to reduce proliferation and angiogenesis. The purpose of this study is to test whether ZD1839 alone will prevent mammary tumors in a mouse model of estrogen receptor negative breast cancer that overexpresses HER-2/neu; and whether ZD 1839 in combination with the COX-2 inhibitor celecoxib will be more effective than either agent alone. The specific aims are: 1. To demonstrate that ZD1839 will reduce incidence, multiplicity, and metastasis of mammary tumors in HER-2/neu mice. 2. To assess whether ZD 1839 will interact with celecoxib, i.e., whether the combination treatment will achieve a reduction in the incidence of mammary tumors in HER-2/neu mice above and beyond what would be expected by the sum of the individual main effects of the two agents. 3. To identify changes in biomarkers associated with the protective effects of the combination ofZD1839 plus celecoxib against mouse mammary tumorigenesis. Specifically, we will identify biomarkers that: (a) are affected by the different treatments; and (b) are associated with the presence/absence of tumors. This study will provide evidence to support the use of the combined treatment of Iressa plus celecoxib for prevention of breast cancer. These findings may also help define biomarker genes associated with the protective effects of the treatments, which may ultimately be used to track response to treatment.